Working Results2
Kaushik Narasimhan
Sys.Date()
Visualization of visitation and dispersal trends against phenology
NFM is the number of mature fruits on a tree. The levels are as follows: a: 1- 50 b: 51 - 200 c: 201 - 500 d: 501 - 1000 e: 1001 - 5000 f: 5000+
The below charts are the NFM compared with real visitation. The X axis is each “tree month” that had a camera in it. The X axis is in sequential order from earliest to latest. It has both high and low cameras. 
##
## Call:
## glm(formula = real.visitation ~ NFM + Height, family = poisson,
## data = cam_data_FG)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -11.357 -6.055 -2.825 1.577 23.117
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 2.96847 0.05943 49.946 < 2e-16 ***
## NFMa 0.06093 0.17021 0.358 0.720
## NFMb -0.52900 0.11930 -4.434 9.24e-06 ***
## NFMc 0.58985 0.05645 10.450 < 2e-16 ***
## NFMd -0.04165 0.05732 -0.727 0.467
## NFMe 0.05648 0.05886 0.960 0.337
## Heightlow 0.60818 0.03128 19.445 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for poisson family taken to be 1)
##
## Null deviance: 7408.9 on 161 degrees of freedom
## Residual deviance: 6429.2 on 155 degrees of freedom
## AIC: 7125.8
##
## Number of Fisher Scoring iterations: 6This doesnt reveal too much in the way of relationships. Lets extract trees that occur across consecutive months.
##
## Call:
## glm(formula = real.visitation ~ NFM + Height, family = poisson,
## data = conmonths)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -12.370 -5.305 -2.507 2.749 17.937
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.85294 0.26124 3.265 0.00109 **
## NFMb 1.98027 0.31250 6.337 2.35e-10 ***
## NFMc 2.72802 0.26013 10.487 < 2e-16 ***
## NFMd 1.79116 0.26081 6.868 6.53e-12 ***
## NFMe 2.23085 0.26008 8.577 < 2e-16 ***
## Heightlow 0.75650 0.03975 19.030 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for poisson family taken to be 1)
##
## Null deviance: 4499.6 on 95 degrees of freedom
## Residual deviance: 3478.9 on 90 degrees of freedom
## AIC: 3884.4
##
## Number of Fisher Scoring iterations: 6
Histograms of 50m bands and 100m
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.00 4.00 19.50 34.28 41.00 237.00## [1] 1962.388## Best Normalizing transformation with 162 Observations
## Estimated Normality Statistics (Pearson P / df, lower => more normal):
## - No transform: 3.4747
## - Log_b(x+a): 3.6518
## - sqrt(x+a): 1.3053
## - exp(x): 5.7466
## - arcsinh(x): 1.531
## - Yeo-Johnson: 1.1246
## - orderNorm: 1.2957
## Estimation method: Out-of-sample via CV with 10 folds and 5 repeats
##
## Based off these, bestNormalize chose:
## Standardized Yeo-Johnson Transformation with 162 nonmissing obs.:
## Estimated statistics:
## - lambda = 0.1337981
## - mean (before standardization) = 3.458438
## - sd (before standardization) = 2.1354313d scatter
Scatter plots using 50m and 100m as is
Scatter of Disperser visitation at 50m and 100m fruiting neighborhood
Scatter of Disperser visitation at 50m and 100m fruiting neighborhood at high cameras
Scatter of Dispersal at low cameras
Scatter plots using 50m and 100m but binarized to presence/absence
Scatter of Disperser visitation at 50m and 100m fruiting neighborhood
Scatter of Disperser visitation at 50m and 100m fruiting neighborhood at high cameras
Scatter of Dispersal at low cameras
Models 50 m band (no data messing)
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: poisson ( log )
## Formula: real.visitation ~ focalmonth..50 + (1 + focalmonth..50 | Tree)
## Data: cam_data_FG
##
## AIC BIC logLik deviance df.resid
## 4515.4 4530.9 -2252.7 4505.4 157
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -8.6574 -2.8417 -0.6314 2.0048 24.2336
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Tree (Intercept) 1.882 1.372
## focalmonth..50 1.725 1.314 -0.90
## Number of obs: 162, groups: Tree, 49
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 2.8575 0.2163 13.213 <2e-16 ***
## focalmonth..50 0.5054 0.2460 2.055 0.0399 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## fclmnth..50 -0.757## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: poisson ( log )
## Formula: real.visitation ~ focalmonth..50 + (1 + focalmonth..50 | Tree) +
## (1 + focalmonth..50 | date)
## Data: cam_data_FG
##
## AIC BIC logLik deviance df.resid
## 3094.9 3119.6 -1539.4 3078.9 154
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -8.6697 -2.1365 -0.1654 1.8407 11.6675
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Tree (Intercept) 1.7878 1.3371
## focalmonth..50 1.0429 1.0212 -0.91
## date (Intercept) 0.9082 0.9530
## focalmonth..50 0.3200 0.5657 -0.30
## Number of obs: 162, groups: Tree, 49; date, 22
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 2.8177 0.2923 9.641 <2e-16 ***
## focalmonth..50 0.6179 0.2570 2.404 0.0162 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## fclmnth..50 -0.523## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: poisson ( log )
## Formula:
## real.visitation ~ focalmonth..50 + CANOPY_DENS_CENSUS_1 + ALTURA_CENSUS_1 +
## CERCROPIA_CENSUS_1 + ARB_DAP50_CENSUS_1 + (1 + focalmonth..50 |
## Tree) + (1 + focalmonth..50 | date)
## Data: cam_data_FG
##
## AIC BIC logLik deviance df.resid
## 3082 3119 -1529 3058 149
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -8.659 -2.155 -0.188 1.863 11.682
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Tree (Intercept) 1.5065 1.2274
## focalmonth..50 0.9978 0.9989 -0.91
## date (Intercept) 0.9216 0.9600
## focalmonth..50 0.3294 0.5740 -0.32
## Number of obs: 161, groups: Tree, 48; date, 22
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 2.7505826 0.5796187 4.746 2.08e-06 ***
## focalmonth..50 0.7021152 0.2621845 2.678 0.00741 **
## CANOPY_DENS_CENSUS_1 -0.0009985 0.0113544 -0.088 0.92993
## ALTURA_CENSUS_1 0.0076531 0.0388320 0.197 0.84376
## CERCROPIA_CENSUS_1 -0.1331298 0.2088527 -0.637 0.52384
## ARB_DAP50_CENSUS_1 -0.0646910 0.2671818 -0.242 0.80868
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) fc..50 CANOPY ALTURA CERCRO
## fclmnth..50 -0.344
## CANOPY_DENS -0.382 0.023
## ALTURA_CENS -0.134 0.056 -0.803
## CERCROPIA_C 0.097 -0.114 -0.252 0.040
## ARB_DAP50_C -0.024 0.086 0.069 -0.176 0.079
## convergence code: 0
## Model failed to converge with max|grad| = 0.00412225 (tol = 0.001, component 1)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: poisson ( log )
## Formula:
## real.visitation ~ focalmonth..50 * Height + (1 + focalmonth..50 |
## Tree) + (1 + focalmonth..50 | date)
## Data: cam_data_FG
##
## AIC BIC logLik deviance df.resid
## 2578.9 2609.8 -1279.5 2558.9 152
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -6.7081 -1.9494 -0.2202 1.2513 10.3249
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Tree (Intercept) 1.7308 1.3156
## focalmonth..50 1.0295 1.0146 -0.86
## date (Intercept) 0.8376 0.9152
## focalmonth..50 0.3490 0.5908 -0.17
## Number of obs: 162, groups: Tree, 49; date, 22
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 2.12546 0.28831 7.372 1.68e-13 ***
## focalmonth..50 0.75897 0.27759 2.734 0.00625 **
## Heightlow 0.84768 0.04932 17.186 < 2e-16 ***
## focalmonth..50:Heightlow -0.07948 0.04016 -1.979 0.04779 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) fc..50 Hghtlw
## fclmnth..50 -0.438
## Heightlow -0.143 0.083
## fclmn..50:H 0.096 -0.104 -0.690Models 50 m (binary fruiting neighborhood)
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: poisson ( log )
## Formula: real.visitation ~ focalmonth..50 + (1 + focalmonth..50 | Tree) +
## (1 + focalmonth..50 | date)
## Data: binFN
##
## AIC BIC logLik deviance df.resid
## 3051.4 3076.1 -1517.7 3035.4 154
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -8.6568 -1.8778 -0.2004 1.7662 11.5114
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Tree (Intercept) 1.9968 1.4131
## focalmonth..501 1.8719 1.3682 -0.87
## date (Intercept) 1.1793 1.0860
## focalmonth..501 0.5046 0.7104 -0.45
## Number of obs: 162, groups: Tree, 49; date, 22
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 2.7197 0.3258 8.348 < 2e-16 ***
## focalmonth..501 1.0187 0.3377 3.017 0.00255 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## fclmnt..501 -0.592## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: poisson ( log )
## Formula:
## real.visitation ~ focalmonth..50 + CANOPY_DENS_CENSUS_1 + ALTURA_CENSUS_1 +
## CERCROPIA_CENSUS_1 + ARB_DAP50_CENSUS_1 + (1 + focalmonth..50 |
## Tree) + (1 + focalmonth..50 | date)
## Data: binFN
##
## AIC BIC logLik deviance df.resid
## 3038.3 3075.3 -1507.2 3014.3 149
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -8.6504 -1.9272 -0.1939 1.7866 11.5394
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Tree (Intercept) 1.7281 1.3146
## focalmonth..501 1.6952 1.3020 -0.88
## date (Intercept) 1.2204 1.1047
## focalmonth..501 0.5128 0.7161 -0.46
## Number of obs: 161, groups: Tree, 48; date, 22
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 2.60430 0.70727 3.682 0.000231 ***
## focalmonth..501 1.09725 0.32871 3.338 0.000844 ***
## CANOPY_DENS_CENSUS_1 -0.01072 0.01289 -0.832 0.405415
## ALTURA_CENSUS_1 0.04245 0.04177 1.016 0.309506
## CERCROPIA_CENSUS_1 0.00850 0.20649 0.041 0.967167
## ARB_DAP50_CENSUS_1 -0.05709 0.25107 -0.227 0.820125
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) f..501 CANOPY ALTURA CERCRO
## fclmnt..501 -0.330
## CANOPY_DENS -0.427 0.036
## ALTURA_CENS -0.128 0.003 -0.788
## CERCROPIA_C -0.023 0.004 -0.252 0.173
## ARB_DAP50_C 0.017 0.048 -0.046 -0.048 -0.090## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: poisson ( log )
## Formula:
## real.visitation ~ focalmonth..50 * Height + (1 + focalmonth..50 |
## Tree) + (1 + focalmonth..50 | date)
## Data: binFN
##
## AIC BIC logLik deviance df.resid
## 2536.9 2567.8 -1258.4 2516.9 152
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -6.8582 -1.9134 -0.1057 1.3413 10.0765
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## Tree (Intercept) 1.9330 1.3903
## focalmonth..501 1.8479 1.3594 -0.86
## date (Intercept) 0.9929 0.9965
## focalmonth..501 0.5051 0.7107 -0.18
## Number of obs: 162, groups: Tree, 49; date, 22
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 2.101854 0.312790 6.720 1.82e-11 ***
## focalmonth..501 1.145907 0.344713 3.324 0.000887 ***
## Heightlow 0.781382 0.050990 15.324 < 2e-16 ***
## focalmonth..501:Heightlow 0.001214 0.071342 0.017 0.986425
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) fc..501 Hghtlw
## fclmnt..501 -0.513
## Heightlow -0.134 0.126
## fclm..501:H 0.096 -0.159 -0.715